Written by 2023 has witnessed unprecedented growth in the field of Artificial Intelligence (AI), marking a significant turning point in its development. This transformation is particularly evident in the realm of assistive technologies for the visually impaired, as highlighted during the 2023 Sight Tech Global conference held last week. The conference, organized annually since 2020 by the Palo Alto-based Vista Center for the Blind and Visually Impaired, provides a platform for industry leaders and academics to discuss advancements in technologies aiding those with blindness and low vision.
A key highlight of the conference was the collaboration between Be My Eyes, an app-based service connecting blind users to volunteer sighted guides through smartphone cameras, and OpenAI‘s GPT-4. This partnership has led to the creation of Be My AI, an innovative application that has been in open beta since September. Be My AI integrates GPT-4 to offer an unprecedented level of descriptive detail, allowing blind individuals to perform tasks like reading instructions, setting up devices, and viewing photographs or artwork independently, without needing human assistance.
The conference also showcased other organizations leveraging AI to assist the visually impaired. AYES.AI, a New York-based company, introduced an app named OKO that helps blind people cross streets by using a vast database of images of pedestrian crossing signals. Similarly, Envision, a company from the Netherlands, has integrated GPT-4 into its state-of-the-art smart glasses. These glasses enable visually impaired users to understand and interact with their surroundings hands-free, offering a groundbreaking approach to navigating visual environments.
These developments underscore the transformative impact of AI growth in 2023, particularly in enhancing accessibility and independence for the 43 million individuals worldwide who are blind. This evolution in AI technology, especially the launch of GPT-4 in March of 2023, has allowed the public to engage with advanced AI on an unprecedented scale and depth, marking a significant milestone in the journey towards more inclusive technology.
The rapidly evolving landscape of wearable technology, particularly in the realm of glasses and headsets, is generating considerable excitement in the field of accessibility. This was a key focus of a session at the Sight Tech Global conference in which Mike Shebanek, Meta’s Head of Accessibility, shared his insights during an interview titled “Where Will AI Take Accessibility?”
Shebanek discussed the implications of Meta’s latest innovations, including the Meta Quest Three AR/VR headset and the Ray-Ban Meta Smart Glasses. These advancements are poised to revolutionize the way people interact with technology, offering hands-free capabilities for posting, texting, snapping, and uploading photos. This shift is particularly significant for blind and visually impaired users, who stand to benefit from two major advancements.
The first notable shift is the move away from traditional mouse and keyboard interactions, facilitated by wearable technologies. This transition paves the way for built-in voice navigation, which has the potential to overcome many of the accessibility challenges inherent in conventional computing. The second, and perhaps more groundbreaking shift, is the integration of Meta’s AI assistant into wearables like the Ray-Ban Smart Glasses. This feature promises to empower visually impaired users in unprecedented ways, by bringing AI into continuous and direct alignment with the human body, surpassing the capabilities of smartphone-based technologies.
Shebanek’s vision for the future, as shared with Sight Tech Global, envisions AI becoming an omnipresent yet unobtrusive companion, seamlessly integrated into our daily lives through devices like glasses or wrist wear. This integration represents a significant stride towards more intuitive and accessible technology interfaces.
Beyond the realm of language models and chatbots, AI’s role in automating tasks and enhancing efficiency cannot be understated, especially for visually impaired individuals. A prime example of this is in the domain of autonomous vehicles. Alphabet’s Waymo has been at the forefront of developing an autonomous taxi service, a development that holds immense promise for individuals with sight loss. This innovation addresses the challenges faced by visually impaired people when using traditional ride-share services, including negative attitudes, safety concerns, and accessibility barriers, offering a glimpse into a more inclusive and accessible future.
This advancement of autonomous vehicle technology by companies like Waymo represents a significant leap forward in accessibility for individuals with visual impairments. Currently operational in cities including Los Angeles, metro Phoenix, Austin, Texas, and recently San Francisco, Waymo’s progress has been bolstered by the advocacy of organizations representing the blind community. In August, the company secured an operating license in San Francisco, marking a notable milestone in the journey towards more accessible transportation solutions.
Tim Elder, President of the National Federation of the Blind of California and an individual living with sight loss, shared his experience of riding in a driverless vehicle at the Sight Tech Global conference. He described the profound sense of independence and autonomy he felt, emphasizing the transformative impact such technology can have on individuals who are visually impaired. Elder’s experience, traveling down a busy San Francisco street with no one in the driver’s seat, highlighted the rapid advancement and growing commonplace nature of this technology, underscoring its potential to truly revolutionize personal mobility.
The conference also delved into the ethical and practical challenges posed by the integration of AI in accessibility. A notable discussion arose around the potential obsolescence of traditional web accessibility practices in the face of evolving AI technologies. Karthik Mahadevan, co-founder of Envision, sparked a debate on whether the future of web accessibility will shift from a user-centered design to an AI-centered approach. The question arises: if AI, through devices like smart glasses, can interpret web content for blind users, is there still a need for websites to be coded for compatibility with human-operated screen reading software?
This intersection of AI and accessibility raises complex questions, which prompted Sight Tech Global to dedicate a session to exploring these issues. One particularly nuanced challenge is the balance between accessibility and privacy. For visually impaired individuals, the desire to recognize faces as sighted people do naturally presents a dilemma. How can digital face recognition be achieved without resorting to invasive methods like a cloud-based facial recognition database? These concerns underscore the intricate balance that must be struck as technology advances, ensuring that the benefits of enhanced accessibility do not come at the cost of privacy or ethical integrity.
During the Sight Tech Global session “AI Gets Complicated: Emerging Debates And Surprising Challenges”, Chancey Fleet, the Assistive Technology Coordinator for the New York Public Library, shed light on the intricate tension between the benefits of facial recognition technology for accessibility and its potential implications for personal privacy. Fleet articulated the complex intersection of providing equitable access to those with visual impairments while also respecting the privacy rights of others. She emphasized the need for ongoing dialogue and collaboration between accessibility advocates and social justice ethicists to navigate this challenging landscape.
Fleet expressed her concerns about the accelerating use of facial recognition for surveillance and tracking purposes, highlighting the dilemma between her own ‘right of access’ and the rights of others who wish to remain unnoticed or anonymous in the digital realm. This tension underscores the broader ethical considerations that must be addressed as AI technologies advance.
In addition, Bryan Bashin, Vice Chair of Be My Eyes, speaking in the same session, offered a perspective grounded in realism. He underscored that activists advocating for the visually impaired are unlikely to support the development of an advanced AI public surveillance network, even if it could potentially improve accessibility. Bashin’s stance reflects a balanced approach, recognizing the potential of AI to significantly enhance and enrich lives, while also acknowledging the importance of implementing these technologies responsibly.
He stressed the necessity for AI developers and thinkers to be aware of the existence and needs of the visually impaired community, ensuring their inclusion in the development process. Bashin’s call for the visually impaired to have a seat at the ‘digital table’ underlines the need for inclusive and responsible AI development, ensuring that technological advancements serve to empower and support all members of society, without compromising individual rights and freedoms.